Search

Speech feature learning is the key to speech mental health recognition. Deep feature learning can automatically extract the speech features but suffers from the small sample problem. The traditional feature extract method is effective, but cannot find the inter-feature structure to generate the new high-quality features. This paper proposes an embedded hybrid feature deep sparse stacked autoencoder ensemble method to solve this problem. Firstly, the speech features are extracted based on prior knowledge and called original features. Secondly, the original features are embedded into the deep network (Sparse Stacked Autoencoder) to filter the output of the hidden layer, to enhance the complementarity between the deep features and the original features. Thirdly, the L1 regularized feature selection mechanism is designed to reduce the hybrid feature set formed by the combination of deep features and original features. Finally, a manifold projection classifier ensemble is designed to enhance the stability of classification. Besides, this paper for the first time proposes a speech collection scheme for mental health recognition. We construct a large-scale Chinese mental health speech database for verification of the proposed algorithm of mental health. In the experimental section, the proposed algorithm is verified and compared with the representative related algorithms. The experimental results show that the proposed algorithm has better classification accuracy than the other representative algorithms. The proposed method combines the advantages of deep feature learning and traditional feature extraction methods more efficiently to solve the small sample problem.

In the SAR change detection algorithm based on self-supervised learning, speckle noise reduces the difference image (DI) quality. Therefore, the contrast of the DI is low, and its change area is not significant. Moreover, the preclassification algorithm with the poor robustness makes the classification results of the low-quality DI inaccurate. When the wrong labels are sent into the classification network, the accuracy of the final detection results is reduced. First, to improve the quality of the initial DI, we design an adaptive gamma correction algorithm that adjusts the contrast according to the mean value of the initial DI and the variation coefficient $\beta$ . The contrast of the new DI generated by this algorithm is higher. Furthermore, to suppress the noise, we adopt a new algorithm based on popular ranking to obtain the saliency map of the new DI. Combining the initial DI with this saliency map, a high-quality DI with a low noise level is obtained. After that, we introduce the structure tensor into the fuzzy local information c-means clustering algorithm (FLICM) to classify the DI more accurately. The new clustering algorithm improves the accuracy of preclassification, especially its hierarchical version. Besides, we use the structure tensor to generate the structure maps of the original images. Finally, according to the prior information obtained from the preclassification, we use a convolution wavelet neural network (CWNN) to supervise and train the structure maps of the original images. Experimental results show that the DI generated by us is closer to the ground-truth than other methods. Our preclassification algorithm performs better. Our algorithm shows higher detection accuracy for SAR images with strong noise than some advanced change detection algorithms.

The diversity of network attacks poses severe challenges to intrusion detection systems (IDSs). Traditional attack recognition methods usually adopt mining data associations to identify anomalies, which has the disadvantages of a high false alarm rate (FAR), low recognition accuracy (ACC) and poor generalization ability. To ameliorate the comprehensive capabilities of IDS and strengthen network security, we propose a novel intrusion detection method based on the adaptive synthetic sampling (ADASYN) algorithm and an improved convolutional neural network (CNN). First, we use the ADASYN method to balance the sample distribution, which can effectively prevent the model from being sensitive to large samples and ignore small samples. Second, the improved CNN is based on the split convolution module (SPC-CNN), which can increase the diversity of features and eliminate the impact of interchannel information redundancy on model training. Then, an AS-CNN model mixed with ADASYN and SPC-CNN is used for intrusion detection tasks. Finally, the standard NSL-KDD dataset is selected to test AS-CNN. The simulation illustrates that the accuracy is 4.60% and 2.79% higher than that of the traditional CNN and RNN models, and the detection rate (DR) increased by 11.34% and 10.27%, respectively. Additionally, the FAR decreased by 15.58% and 14.57%, respectively, compared with the two models.

This paper studies fuzzy adaptive finite-time stabilization control issue of uncertain stochastic nonlinear high-order systems. Fuzzy logic system (FLS) is used to identify unknown nonlinearities. The unknown control direction is handled by adopting the Nussbaum function. Combining adaptive backstepping design with adding a power integrator method, a fuzzy adaptive finite-time control approach is developed, which guarantees that the closed-loop system is semi-global practical finite-time stable in probability.

Cross-modal retrieval is a very challenging and significant task in intelligent understanding. Researchers have tried to capture modal semantic information through a weighted attention mechanism. Still, they cannot eliminate irrelevant semantic information's negative effects and cannot capture fine-grained modal semantic information. In order to further accurately capture the multi-modal semantic information, a bidirectional focused semantic alignment attention network (BFSAAN) is proposed to handle cross-modal retrieval tasks. Core ideas of BFSAAN are as follows: 1) Bidirectional focused attention mechanism is adopted to share modal semantic information, further eliminating the negative influence of irrelevant semantic information. 2) Strip pooling is applied to image and text modalities, a lightweight spatial attention mechanism to capture modal spatial semantic information. 3) Second-order covariance pooling is explored to obtain multi-modal semantic representation, capturing modal channel semantic information and achieving semantic alignment between image-text modalities. The experiment is executed in two standard cross-modal retrieval datasets (Flickr30K and MS COCO). The experimental design includes four aspects: performance comparison, ablation analysis, algorithm convergence, and visual analysis. Experimental results show that BFSAAN has better crossmodal retrieval performance.

This paper investigates the problem of adaptive fuzzy 1-bit event-triggered control for stochastic nonlinear systems. In our design, we only need to transmit a 1-bit signal (either 1 or 0), whenever the triggering signal is updated. That is, the signal is transmitted between the controller and the actuator in the communication channel. With our proposed control scheme, which can further alleviate the communication burden. Based on the Lyapunov stability analysis, all the closed-loop signals are semi-globally uniformly ultimately bounded (SGUUB). Meanwhile, Zeno behavior can be avoided.

Electrical conductors and electrical equipment were selected according to the mean of annual lowest temperature, and it was of high likelihood to suffer abnormal lower temperature than its designed criterion. Sealing component of Gas Insulated switchgear (GIS) should be designed based on the criterion of minimum temperature, which could avoid the risk of large-area SF6 leakage accidents causing by ultra-design low temperature. Nine stations in Shandong Province were selected as the research objects, and the temperature observation data from 1973 to 2016 were investigated in the manuscript. The risk level of extreme temperature lower than the design criterion was calculated with Generalized Extreme Value (GEV) distribution according to the annual minimum temperatures determined by the current regulations. Furthermore, the risk of encountering ultra-design low temperature were also evaluated based on the local historical observation minimum temperature and the minimum design temperature in typical areas, which could provide theoretical support and application reference for optimizing the design temperature of power transmission and transformation equipment.

In this paper, an adaptive fuzzy control scheme is proposed for one-quarter automotive active suspension system with full sate constraints and stochastic disturbance. In the considered active suspension system, to further improve the driving security and comfort, the problems of stochastic perturbation and full state constraints are considered simultaneously. In the framework of backstepping, the barrier Lyapunov function is proposed to constrain full state variables. Consequently, by combing the Ito differential formula and stochastic control theory, an adaptive controller is designed to adopt the uneven pavement surface. Ultimately, on the basis of Lyapunov stability theory, it proves that the designed controller not only can constrain the bodywork, the displacement of tires, the current of the electromagnetic actuator, the speeds of the car body and the tires within boundaries, but also can eliminate the stochastic disturbance.

In economic growth theory, force of long-run economic growth is considered to be capital accumulation, exogenous then endogenous technology progress. Economic aggregate cannot be represented as a production function of only production factors because it is related to both demand and supply. In this study, economic aggregate is divided into business revenue of growing product and that of mature product. Growing product provides surplus demand and mature product consumes surplus demand. Economic aggregate depends on innovation, the amount of growing products. In addition, two kinds of innovation, invention of new consuming products and technology progress play different role in economic growth. The former brings final demand increase and the latter leads to production cost reduction. Technology progress often exceeds new invention of consuming product because of strong externality.

This paper addresses the adaptive fuzzy finitetime control problem for switched nonlinear inverted pendulum systems. The considered inverted pendulum systems are switched model between different springs with under actuated and strongly coupled characteristics. A finite-time controller is constructed by using adding one power integrator technique, adaptive fuzzy control method and backstepping scheme. The proposed control approach is capable of ensuring the semi-global practical finite-time stability for the specific mechanical systems.

Focused Attention Level (FAL) is important in many tasks such as driving, online learning, and many computer-related operations. Previous approaches to evaluation of FAL usually involve eye-tracking and electroencephalogram which require dedicated devices to track the movements of eyes or record the signals from the brain. In this paper, we propose a novel method of evaluation of FAL only using a user's mouse operation data which are easily available in computers equipped with a mouse device. We perform our study using the Schulte Table which is a popular task designed to improve the focused attention levels of children and students. We extract features from mouse operation data in completing a Schulte Table task and establish a machine learning model to predict the focused attention level of the subject. Experimental results show that the accuracy of our method can reach 80.9%.

This paper proposes a novel adaptive sampled-data control strategy subject to a category of multiply-input multiply-output (MIMO) system with the pure feedback form. The unknown functions connect each subsystem of the controlled plant, the states of the original plant are unavailable, and the output variable is the only accessible data of the original system. Fuzzy logical systems are implemented to address the uncertainties, and the immeasurable states are estimated by designing a novel sampled-data observer. On the basis of the framework of backstepping, an output feedback sampled-data controller is designed. The proposed virtual control signals and sampled-data controller are composed of the sampled-data. According to the Lyapunov theory and Gronwall-Bellman inequality, it proves that the SGUB property of all signals in the original system.

This paper focuses on the issue of fuzzy adaptive control design for switched high-order nonlinear systems with time-varying output constraints. Fuzzy logic systems (FLSs) are adopted to identify the unknown nonlinearities. To handle the time-varying output constraint, a newly asymmetric timevarying barrier function is considered in control design. By adopting adaptive backstepping control technique, a fuzzy adaptive controller is designed based on average dwell time (ADT) approach. It means that all signals of the considered system are semi-globally uniformly ultimately bounded (SGUUB) and the asymmetric yet time-varying output constraint does not beyond.

In this paper, the finite-time fault-tolerant control design problem is addressed for non-strict feedback nonlinear systems. The considered nonlinear systems contain bias fault and loss of effectiveness in actuator, which can be addressed by fault-tolerant control technique. The fuzzy logic systems (FLSs) are introduced to approximate the unknown nonlinear functions defined on a compact set. Meanwhile, the technique named adding a power integrator is used, which validly overcomes the problem of the system with power-exponentials. A finite-time fault-tolerant controller is constructed base on backstepping design technique and practical finite-time stability theory. Finally, validity is illustrated by the finite-time fault-tolerant control strategy and a simulation.

This work addresses the finite-time control problem for a class of nonlinear inverted pendulum systems. A series of homogeneous controllers, which are capable of guaranteeing the locally finite-time stability for the closed-loop systems, are first developed using the adding one power integrator method and backstespping technique. Subsequently, the nested saturation control approach has been further proposed in order to achieve global finite-time stability after appropriately adjusting the saturation level. Furthermore, the simulation results are given to validate the effectiveness of the proposed strategy.

An event-trigger control approach of nonlinear systems is presented. The saturated controllers are given by applying event-trigger condition. Using Lyapunov-Krasovskii functional technique, we prove the asymptotically stability of fuzzy systems. The controllers are obtained by the matrix inequalities. Finally, the method is substantiated with numerical example.

This work studies an observer-based fuzzy adaptive control problem for the uncertain nonlinear time-delay systems with unknown virtual and actual control gain functions. The Lyapunov-Krasovskii function is utilized to eliminate the unknown time delays. In order to estimate the uncertain nonlinear functions, Fuzzy Logic Systems(FLSs) are quoted. Then the fuzzy state observer is devised to handle the unavailable state issue. By using the backstepping control technique and bounded control method, a novel observer-based fuzzy adaptive backstepping control approach is developed. The rationality of the presented control methods is demonstrated by means of the Logarithm Lyapunov functions.

Efficiency improvement is the core element in state-owned assets management of colleges and universities. Internet thinking and technology create favorable conditions for efficiency improvement of assets management. In this study, state-owned assets data of colleges and universities were analyzed and typical problems against efficiency improvement of asset management was analyzed. Finally, strategies for efficiency improvement of asset management of colleges and universities based on Internet thinking were proposed.

For Takagi-Sugeno fuzzy uncertain continuous systems, this paper presents event trigger control. First, the fuzzy controllers are designed. Second, by applying of Lyapunov function theory, stabilization of the fuzzy uncertain continuous systems is proved by applying linear matrix inequalities theory and average dwell time control law. At last, a simulation example verifies the correctness and effectiveness of the approach and conditions.

In this paper, the finite-time optimal control problem is investigated for a class of nonlinear lower-triangular systems. In the control design process, based on the Lyapunov function and adding a power integrator method, a control law is designed to make the nonlinear lower-triangular systems locally finite-time stability. Under the nested saturation control technique, the nonlinear lower-triangular systems is globally finite-time stabilization by adjusting the saturation. Then, according to the basic idea of optimization, the proposed control method can make the cost function be minimized by selecting the appropriate parameters in the controller, and achieves the goal of control optimality. Finally, a simulation example is provided to show the effectiveness of the presented control method.

This study develops an adaptive neural inverse optimal control method for a class of stochastic nonlinear systems. Neural networks (NN) are used to approximate the unknown nonlinear functions. The designed inverse optimal control strategy avoids the objective of solving the Hamilton-Jacobi-Bellman (HJB) equation and devises an optimal controller, which is related to the meaningful cost functional. Based on adaptive backstepping algorithm and Lyapunov stability theory, it is proved that the proposed control strategy guarantees the asymptotic stability in probability of the control systems and solves the inverse optimal problem.

The finite element method (FEM) based on unstructured grid has been widely used in science and engineering. Charm++ programming has the advantages of task over-decomposition, asynchronous call communication and dynamic load, and is very suitable for parallelization of finite element methods which is unstructured grid discrete. This paper designed the Charm++ parallel algorithm of unstructured grid partition based on METIS (serial graph partitioning and fill-reducing matrix ordering). We implement the preprocessing communication modules use adding virtual boundary elements in order to minimize the amount the data for communication. Using insertion sorting and local-to-global element information can ensure the consistency of element id between serial and parallel program. The parallel grid partition can support FEM with unstructured grid in scientific and engineering applications. The testing shows that the partition algorithm is accurate and efficient.

This study develops an adaptive fuzzy inverse optimal control method for a class of two order nonlinear systems with unknown nonlinear dynamics. By using the knowledge of a control Lyapunov function and the universal approximation property of a fuzzy logic system, the inverse optimal control design and unknown nonlinear dynamics problems are solved, respectively. A new adaptive fuzzy optimal controller is constructed based on backstepping design principle, and the stability and optimality are proven by using rigorous mathematical reasoning.

A fuzzy adaptive control method is developed for active suspension systems with unknown dynamics. The controlled object is a single-input multiple-output five-order active suspension system with electromagnetic actuator. Fuzzy logic systems (FLSs) are utilized to identify the unknown nonlinear dynamics. Based on the back-stepping design principle and Lyapunov stability theory, the proposed control scheme is proved to guarantee that the closed-loop system is stable. Simulation results are presented to demonstrate the effectiveness of the developed control strategy.

In this paper, an observer-based adaptive fuzzy backstepping control approach is proposed for a class of uncertain nonlinear stochastic systems with the immeasurable states and unknown-time-delays. The fuzzy logic systems are used to approximate the unknown nonlinear functions, and a fuzzy state observer is designed to estimate the unmeasured states. By combining the backstepping design technique, an observer-based adaptive fuzzy control method is developed. It is proved that the proposed adaptive fuzzy control approach can not only guarantee that all the signals of the closed-loop system are semi-globally uniformly ultimately bounded (SGUUB) in mean square, but also the output of the system converge to a small neighborhood of the origin.

In this paper, the problem of active fault-tolerant control (FTC) is investigated for multi-agent systems with unknown control direction and actuator faults. The communication topology among the agents is directed and exist one leader that has a directed path to each follower at least. The system considered in this paper is consist of unstructured uncertainties, unknown control direction, and actuator faults. The fuzzy logic system is used for approach unknown nonlinear function, the problem of unknown control direction can be solved by using Nussbaum gain function. Based on backstepping design principle and the dynamic surface control technique, a fault-tolerant containment control scheme is developed. By using Lyapunov function stability theory, the proposed FTC controller can guarantee that all signals in all subsystems are bounded, and the tracking errors for each subsystem converge to a small neighborhood of zero.

This paper studies the set operations on maxitive belief structures. Unlike Dempster-Shafer structures, we show that different maxitive belief structures can induce the same upper and lower possibilities. Then the operations on the set of maxitive belief structures might not be the operations on their induced upper and lower possibilities. We call those operations on maxitive belief structures which are also the operations on their induced upper and lower possibilities the nice operations. We introduce the set operations on maxitive belief structures, including weighted sum, intersection, union, complement, projection, cylindrical extension and Cartesian product operations. We show that the weighted sum, union, projection and cylindrical extension are nice operations, and the intersection, complement and Cartesian product operations are not nice.

This paper gives an intuitionistic interpretation of maxitive belief structure. Central to this interpretation is that we can use maxitive belief structure to convey correct information for any intuitionistic statement which is a pair of values, (L(A), D(A)), providing information on the value of a variable lying in the set A. Then we can represent an intuitionistic statement by a maxitive belief structure. In fact, we construct the most appropriate representation of an intuitionistic statement based on the entailment of maxitive belief structures. Using this type representation, we introduce methods to combine intu-itionistic statements using the combining rule of maxitive belief structures and making inferences from these type propositions using maxitive belief structures.

Speech feature learning is very important for the design of classification algorithm of Parkinson's disease (PD). Existing speech feature learning method for classification of PD just pays attention to the speech feature. This paper proposed a novel hybrid feature learning algorithm which puts the features of all the speech segments of each subject together, thereby obtaining new and high efficient features without feature transformation. Firstly, hybrid features was constructed by combining features and segments. Secondly, high efficient hybrid feature selection was conducted by various criteria. Thirdly, the selected hybrid features were applied for classification of PD. Besides, various evaluation criteria are introduced into feature selection in this manuscript. Experimental results show that this proposed algorithm can obtain new features (hybrid feature) with satisfactory classification accuracy. The selected features are very stable and meaningful.

Received Signal Strength Indicator (RSSI) is simplest among various indoor localization methods. The main drawback is requiring a training phase to build fingerprints database, which is time consuming and sensitive to changes of test environment. Time based approaches such as Time of Arrival (ToA) and Time Difference of Arrival (TDoA) achieve better performance with some hardware expenditure, while location accuracy is significantly affected by multipath interference. By employing Super-resolution techniques (SRTs), multipath components can be estimated to improve localization precision, but limited to OFDM-based Multiple Input Multiple output (MIMO) wireless communication systems. Considering for general radio signals with strong correlation, we propose an enhanced TDoA approach based on RSS-assisted cross-correlation method. Simulation and experimental results demonstrate that our method is effective to mitigate multipath interference, verified by IEEE 802.11b WLAN. Compared with SRTs, our method achieves comparable localization precision, and has superiority in computational complexity and anti-noise performance.

In this paper, the problem of fault-tolerant control (FTC) is investigated for a class of nonlinear single input and single output (SISO) systems in the non-strict feedback form. The considered system possess unknown nonlinear functions, measured states, unknown control direction and actuator faults (bias and gain faults). With the help of Neural Network System (NNs) to approximate the unknown nonlinear functions and a fuzzy adaptive observer is designed to estimate the unmeasured states. In the frame of adaptive backstepping design technique, and by combining with the nonlinear fault-tolerant control theory, the control performance of the closed-loop system is guaranteed by suitably choosing the design parameters, an adaptive NNs output feedback FTC method is developed. It is shown that the given control approach can prove that all signals in the closed-loop system are proved to be bounded, and the tracking errors between the system outputs and the reference signals converge to a small neighborhood of zero by choice of the design parameters.

The issue of the adaptive dynamic programming control design is explored for time-delay nonlinear systems. The established scheme is constituted of two parts, feedforward control and feedback optimal control, respectively. By employing the universal approximation character of fuzzy logic system (FLS) to handle unknown utility function, thus an optimal fuzzy adaptive feedback design is structured. The stability of the controlled system is further approved, and the output can track the desired trajectory in an optimal manner. The feasibility of the established control tactics is confirmed via simulation case.

In this paper, an output-feedback fuzzy adaptive fault tolerant control design method is presented for a class of nonlinear systems having unmeasured states and actuator faults (bias and gain faults). Firstly, fuzzy logic systems (FLS) are adopted to identify the unknown nonlinearity, and the hyperbolic tangent function is employed as the robust term to overcome the actuator faults. Then, the threshold strategy about tracking error is constructed as system fault alarm. By use of the adaptive backstepping modular scheme, a robust fuzzy adaptive fault tolerant control strategy is presented. It is shown that all the variables in all subsystems are bounded, and also the tracking error converges to a small neighborhood of zero. Finally, the applicability of the proposed controller is well carried out by a simulation example.

Model checking of linear-time properties based on possibility measures was developed by Li and Li (2013). However, the linear-time properties considered in the previous work were classical and qualitative, possibility information of the systems was not considered thoroughly at all. Therefore, the quantitative model checking of fuzzy linear-time properties based on generalized possibility measures was studied by Li (2016). This paper is a continuation of the above work. In this paper, we study the expressive power of possibilistic linear-temporal logic (PoLTL). Unlike possibilistic computation tree logic (PoCTL) in which the expressiveness of PoCTL is more powerful than computation tree logic (CTL), it is surprising that PoLTL is as expressive as linear-temporal logic (LTL). Furthermore, the comparison between PoLTL and generalized PoLTL (GPoLTL) is given. It is shown that the expressiveness of GPoLTL with restrained conditions is the same as PoLTL when semantics of GPoLTL is interpreted in the frame of possibilistic Kripke structure.

In this paper, the mechanism of fault diagnosis and fault warning is set up for a class of single-input single-output (SISO) nonlinear systems with actuator faults, and based on the established fault diagnosis mechanism, an adaptive fuzzy active fault-tolerant control (FTC) strategy is proposed. The considered actuator faults are bias and gain faults. The fault warning is constructed according to the identification error came from controlled system and fault diagnosis system. When the actuator faults occur, fuzzy logic systems are used to identified the unknown nonlinear function, and the adaptive control method and the symbolic function technique are employed to compensate the bias and gain faults, respectively. The stability of the closed-loop system is rigorously investigated. Finally, simulation results of a numerical example are presented to further demonstrate the efficiency of the proposed control strategy.